https://github.com/manwapastorelli/OpenSimLSL

you should find all those issues are resolved at this point. 

Reads a notecard with a supplied name. 
It will ignore all lines that are blank
It will ignore all lines which start with a #
and any lines with do not contain an = sign

Then it will clean up the bits either side of the equals sign to try and account for user inconsistencies. 

If you are reading numbers, vectors or rotations don't forget to combine these with the mehods in my last few threads to validate the user provided you with a valid entry!

ProcessInstructionLine(string instruction, string data)
{
    /*
    This is where you add code to process the instuction based on the notecard content. 
    For example if this is a notecard containing a list of names who need to be on an access list
    you might have a line like:
    if (instruction == "name") accessList += data;
    */
}

ReadConfigCards(string notecardName)
{   // reads the config cards and restores the settings
    if (llGetInventoryType(notecardName) == INVENTORY_NOTECARD)
    {   //come here if the requested notecard exists in the inventory
        integer notecardLength = osGetNumberOfNotecardLines(notecardName);
        integer index;
        for (index = 0; index < notecardLength; ++index)
        {
           string currentLine = osGetNotecardLine(notecardName,index);
           string firstChar = llGetSubString(currentLine, 0,0);
           integer equalsIndex = llSubStringIndex(currentLine, "=");
           if (currentLine != "" && firstChar != "#" && equalsIndex != -1 )
           {
                //come here if the line is not blank AND
                //the first character is not "#" AND
                //the line contains an  "=" sign.
                string instruction = llGetSubString (currentLine, 0, equalsIndex-1); //everything before the equals sign
                string data = llGetSubString(currentLine, equalsIndex+1, -1); //everyting after the euquals sign
        Â

please note this method relies on you also including another method to check if a string is a float. This can be found in a previous post in this forum

#include "chkisrot.lsl"
integer isRot = ChkIsRot (string inputString)

integer ChkIsRot (string inputString)
{ // returns a true if supplied string is a vector and false if it is not
    /*
    Relies on you including
    chkisfloat.lsl
    */
    integer isRot;
    string startProcess = llStringTrim ((inputString), STRING_TRIM);
    integer stringLength = llStringLength(startProcess);
    integer openBraceIndex = llSubStringIndex(startProcess, "<");
    integer closeBraceIndex = llSubStringIndex(startProcess, ">");
    if (openBraceIndex == 0 && closeBraceIndex == stringLength-1 )
    {
        integer startIndex = openBraceIndex +1;
        integer endIndex = closeBraceIndex -1;
        startProcess = llGetSubString(startProcess, startIndex, endIndex); //removes the first and the last characters ("<" and ">")
        list stringParts = llCSV2List(startProcess); //turn the remaining list elements into a list
        string strX =  llStringTrim( (llList2String(stringParts,0)), STRING_TRIM); // new string for x with whitespace removed
        string strY =  llStringTrim( (llList2String(stringParts,1)), STRING_TRIM); // new string for y with whitespace removed
        string strZ =  llStringTrim( (llList2String(stringParts,2)), STRING_TRIM); // new string for z with whitespace removed
        string strS =  llStringTrim( (llList2String(stringParts,3)), STRING_TRIM); // new string for z with whitespace removed
        integer strXIsFloat =  ChkIsFloat(strX);
        integer strYIsFloat =  ChkIsFloat(strY);
        integer strZIsFloat =  ChkIsFloat(strZ);
        integer strSIsFloat =  ChkIsFloat(strS);
        if (strXIsFloat && strYIsFloat && strZIsFloat && strSIsFloat) isRot  = TRUE;
        else isRot  = FALSE;
    }
    else

Posting code snippets in the forum would be much clearer with a little BB code support. 

Since the quote already works I suspect it would not take much to make the CODE work as well. It already recognizes the tags but the outcome isn't great. It removes all the line returns, which actually makes things worse.